The origins of this book can be traced back to a conference paper presented almost twenty years ago with Giovanni Arrighi at the Seventh Political Economy of the World-System conference organized by Charles Bergquist at Duke University. The paper, entitled “Labor Movements and Capital Mobility: the United States and Western Europe in World-Historical Perspective,” was a first attempt at understanding how outcomes for different national labor movements are linked to each other by world-economic processes, especially the transnational relocation of capital. Over the years, this first seed grew, fused with other seeds, and developed into the present book. Ongoing exchanges with Giovanni Arrighi over this entire time period have left an indelible mark on the final product, and I therefore start by acknowledging my intellectual debt to him.

Another debt is owed to the members of the World Labor Research Group – a group of faculty and graduate students who met regularly at the Fernand Braudel Center at Binghamton University in the 1980s. Apart from myself, the members of the research group were Giovanni Arrighi, Mark Beittel, John Casparis, Jamie Faricellia Dangler, Melvyn Dubofsky, Roberto Patricio Korzeniewicz, Donald Quataert, and Mark Selden. It was in the course of discussions within this group that it became clear that the serious study of labor movements from a global and historical perspective would require new types of data that were simply not available in existing compilations. In 1986 the group plunged into a massive data collection project, initiating the World Labor Group (WLG) database, on which this book stands.

It soon became clear that the creation of this database would require an enormous effort and that it risked never being completed.

This chapter analyzes the world-scale dynamics of labor unrest in what is widely regarded as the leading industry of twentieth-century capitalism – the world automobile industry. The first section of the chapter presents a picture of the temporal and spatial patterning of labor unrest in the world automobile industry from 1930 to the present based on indexes derived from the World Labor Group database. We identify a series of spatial shifts in the distribution of labor unrest – with the epicenter of militancy moving from North America through Western Europe to a group of newly industrializing countries over the course of the twentieth century.

The second section then narrates the dynamics of these spatial shifts and their interrelationship with successive rounds of capital relocation. We argue that mass production in the automobile industry has tended to recreate similar social contradictions wherever it has grown, and, as a result, strong and effective labor movements have emerged in virtually every site where Fordist mass production expanded rapidly. Yet each time a strong labor movement emerged, capitalists relocated production to sites with cheaper and presumably more docile labor, weakening labor movements in the sites of disinvestment but strengthening labor in the new sites of expansion.

Our story of the interrelationship between labor movements and capital relocation thus paints a far more ambiguous picture than that suggested by the race-to-the-bottom thesis (see Chapter 1). To put it in a sentence, the trajectory for the world automobile industry suggests that where capital goes, conflict goes. Or to paraphrase David Harvey (1989: 196; 1999: 390, 442), the geographical relocation of production is a “spatial fix” that only “reschedules crises” it does not permanently resolve them.

Crisis of Labor Movements and Labor Studies

During the last two decades of the twentieth century, there was an almost complete consensus in the social science literature that labor movements were in a general and severe crisis. Declining strike activity and other overt expressions of labor militancy (Screpanti 1987; Shalev 1992), falling union densities (Western 1995; Griffin, McCammon, and Botsko 1990) and shrinking real wages and growing job insecurity (Bluestone and Harrison 1982; Uchitelle and Kleinfeld 1996) were among the trends documented. The bulk of the empirical literature focused on trends in wealthy countries (especially North America and Western Europe), yet many saw the crisis as world-scale, adversely affecting labor and labor movements around the globe.

This sense that labor movements are facing a general and severe crisis contributed to a crisis in the once vibrant field of labor studies. As William Sewell (1993: 15) noted: “Because the organized working class seems less and less likely to perform the liberating role assigned to it in both revolutionary and reformist discourses about labor, the study of working class history has lost some of its urgency” (see also Berlanstein 1993: 5).

For many, this double crisis of labor studies and labor movements is long term and structural – intimately tied to the momentous transformations that have characterized the last decades of the twentieth century going under the general rubric of “globalization.” For some, the crisis is not just severe, it is terminal. Aristide Zolberg, for one, argued that late-twentieth-century transformations have brought about the virtual disappearance of “the distinctive social formation we term ‘working class.’”

This appendix contains the instructions used by persons recording data from newspaper indexes. Examples of how to classify events, the introductory statement, and the lists of index headings to use as a guide in searching the index are not included here. The latter is a several-pages-long list of industries, countries, and other subject headings.

What Types of Reported Actions to Record

1. Record any action indicative of labor unrest (see definition).

2. “Labor” includes wage workers and the unemployed (it does not include peasants, students, soldiers, communists, etc., but it does include agricultural wage laborers). In the case of actions by the unemployed, write “unemployed” in the industry column of the coding sheet.

3. Record actions even if they are only rumored, threatened, or planned, or if the report is that the action is completed and over. Also record actions that have been cancelled.

4. Record whether the act is referred to in an editorial, is a report of persons commenting on the action, or an analysis of the impact of the action (e.g., on the nation's economy).

5. Reports of state action against labor should be recorded in the following way:

1. (a) the index clause mentions a government action only, record it if it is indicative of labor unrest (e.g., anti-strike legislation, arbitration);

2. (b) an index clause mentions both a labor action and a government action, do not record the government action unless:

1. (i) government action indicates a state of siege, coup d'état, or martial law in response to labor unrest (government sending in troops can be interpreted as martial law; arrests should not be);

2. […]

The previous two chapters focused on labor and global economic dynamics – especially the interrelationship between transformations in the organization/location of production, workers' bargaining power, and world-historical patterns of labor unrest. In this chapter, we shift the angle of vision, bringing to center stage the interrelationship between global politics and labor movements. For as we argued in Chapter 1, global economic processes are themselves deeply embedded in global political dynamics ranging from state formation and citizenship bounding to interstate conflict and world war.

In Chapter 1, we also suggested that the twentieth century – viewed through a Polanyian lens – traces a pendulum-like trajectory between the commodification of labor and the breakdown of established social compacts, on the one hand, and the decommodification of labor and the establishment of new social compacts, on the other hand. The first swing of the pendulum – the late-nineteenth- and early-twentieth-centuries movement toward the “commodification of labor” and the initial countermovement response from growing labor movements – is the focus of Section II of this chapter. Section V, in turn, focuses on the pendulum's swing back – the establishment of new national and international social compacts binding labor, capital, and states, which partially protected labor from the vagaries of an unregulated global market in the post–Second World War decades.

This swing of the pendulum was a response to four decades of world war, depression, explosive labor militancy, and worldwide revolutionary upheaval. This intervening period, characterized by a widening and deepening vicious circle of war and labor unrest, is the focus of Sections III and IV.

The World Labor Group (WLG) database is one of the key empirical sources used in this book to document world-historical patterns of labor unrest. This database originates in a collective research effort by a group of graduate students and faculty (The World Labor Research Working Group) at the Fernand Braudel Center (Binghamton University) in the 1980s. The outcome of the group's work was published as a special issue of Review – hereafter referred to as “the special issue” (see Silver, Arrighi, and Dubofsky 1995). The present author subsequently expanded and updated the database produced in the first phase of the project.

This appendix describes the WLG data collection project including issues of conceptualization, measurement, and data collection procedures (see Silver 1995a in the special issue for a more in-depth treatment of these issues). The next section discusses the conceptualization of labor unrest used by the World Labor Group (see also Chapter 1). The second section discusses measurement issues, the third section discusses data collection procedures, and the fourth section discusses the outcome of various reliability studies. Finally, Appendix B reproduces the data collection instructions used for compiling the WLG database.

The Concept of World-Scale Labor Unrest

Efforts to gain an adequate picture of the long-term, world-scale patterns of labor unrest face special problems of conceptualization and measurement. Workers' resistance has taken a variety of forms over the space and time of the world economy. While it might at first seem intuitively obvious, the concept of labor unrest as a world-historical phenomenon and how one might go about measuring it are far from obvious.

Our premise at the start of this book was that, by recasting labor studies in a world-historical framework, we would be able to shed a new light on the contemporary global crisis of labor movements. The central chapters of the book sought to distinguish, from successive angles of vision, dynamics that are recurrent from those that are fundamentally new and unprecedented in the trajectory of world labor unrest. This final chapter returns to the debates highlighted in Chapter 1 about the causes, depth, and nature of the contemporary crisis of labor movements, informed now by our study of the past.

A Race to the Bottom?

Our analysis of the globalization of mass production in the world automobile industry in Chapter 2 concluded that the geographical relocation of production has not created a simple race to the bottom. Rather, we found a recurrent pattern in which the geographical relocation of production tended to create and strengthen new working classes in each favored new site of investment. While multinational capital was attracted by the promise of cheap and controllable labor, the transformations wrought by the expansion of the industry also transformed the balance of class forces. The strong labor movements that emerged succeeded in raising wages, improving working conditions, and strengthening workers' rights. Moreover, they often played a leading role in democracy movements, pushing onto the agenda social transformations that went well beyond those envisioned by pro-democracy elites.

To be sure, the relocation of capital from existing sites of production tended to weaken established working classes.

For Figures 4.2 and 4.3, a very broad rough classification of countries was used. Countries in North America (except Mexico), Europe (both east and west), and Australia and New Zealand were included in the metropolitan aggregate (Figure 4.2). Countries in Asia (east and south), North Africa and the Middle East, Latin America, and Africa were included in the colonial and semi-colonial aggregate (Figure 4.3).

Countries included in the metropolitan aggregate (Figure 4.2) (1990 names; countries with fewer than 100 labor unrest mentions in the WLG database not included in this list): North America (Canada, United States); Europe (Albania, Austria, Belgium, Bulgaria, Cyprus, Czechoslovakia, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Malta, Netherlands, Norway, Poland, Portugal, Romania, Spain, Sweden, Switzerland, United Kingdom, USSR/Russia, Yugoslavia); Oceania (Australia, New Zealand).

Countries included in the colonial and semi-colonial aggregate (Figure 4.3) (1990 names; countries with fewer than 100 labor unrest mentions in the WLG database not included in this list): Middle East and North Africa (Algeria, Egypt, Iran, Israel, Lebanon, Morocco, Sudan, Tunisia, Turkey); Latin America (Argentina, Bolivia, Brazil, Chile, Colombia, Cuba, Dominican Republic, Ecuador, EI Salvador, Guyana, Jamaica, Mexico, Nicaragua, Panama, Peru, Puerto Rico, Trinidad & Tobago, Uruguay, Venezuela); Africa (Ghana, Kenya, Nigeria, South Africa, Zambia, Zimbabwe); Asia (Bangladesh, Burma, China, Hong Kong, India, Japan, Korea, Malaysia, Pakistan, Philippines, Singapore, Sri Lanka).

Countries included in the world aggregate: Figure 4.1 includes all countries with mentions of labor unrest in the WLG database.

In the previous chapter our analysis of world labor unrest focused on the leading capitalist industry of the twentieth century. We followed the rise, globalization, and transformation of the mass production automobile industry. We found a cyclical pattern of labor militancy and capital relocation – a kind of déjà u pattern in which strong labor movements emerged in each new favored low-wage site to which the industry relocated. In other words, spatial fixes re-created similar working classes and class conflict wherever capital went.

In this chapter, we widen the temporal scope of the analysis. On the one hand, we move back in time in order to compare the dynamic of the automobile industry with that of the leading industry of the nineteenth century – the textile industry. On the other hand, we move forward in time in an effort to identify the leading industries of the twenty-first century and to compare their likely dynamics with those of the past.

Two arguments are central to this chapter. The first is that the main location of working-class formation and protest shifts within any single industry along with shifts in the geographical location of production. In other words, we argue that a similar déjà u pattern to that which we found for the world automobile industry can be observed also in the world textile industry. Moreover, we argue that, just as labor unrest shifts from location to location within any given industry, so the main sites of working-class formation and protest shift from industry to industry together with the rise/decline of leading sectors of capitalist development.